Initiation Mechanisms for Radical Polymerization of Methyl

Rosebank MDC, Clayton, Victoria 3169, Australia; and Fine Chemicals and Polymers Research. Laboratory, NOF Corporation, Taketoyo-cho, Chita-gun, Aichi...
1 downloads 0 Views 204KB Size
10824

J. Am. Chem. Soc. 1996, 118, 10824-10828

Initiation Mechanisms for Radical Polymerization of Methyl Methacrylate with tert-Butyl Peroxypivalate Tomoyuki Nakamura,†,§ W. Ken Busfield,† Ian D. Jenkins,*,† Ezio Rizzardo,‡ San H. Thang,‡ and Shuji Suyama§ Contribution from the Faculty of Science and Technology, Griffith UniVersity, Nathan, Queensland 4111, Australia; CSIRO, DiVision of Chemicals and Polymers, PriVate Bag 10, Rosebank MDC, Clayton, Victoria 3169, Australia; and Fine Chemicals and Polymers Research Laboratory, NOF Corporation, Taketoyo-cho, Chita-gun, Aichi 470-23, Japan ReceiVed July 8, 1996X

Abstract: The reaction of tert-butyl peroxypivalate (2) with methyl methacrylate (3) has been studied by the radical trapping technique employing 1,1,3,3-tetramethyl-2,3-dihydro-1H-isoindol-2-yloxyl (1) as a scavenger. Thermolysis of 2 generated tert-butoxyl, tert-butyl, and methyl radicals in the ratios of 48:50:2 at 60 °C in 3. Both alkyl radicals underwent selective tail addition to 3. tert-Butyl radicals reacted about twice as fast as methyl radicals with 3. The absolute rate constant for addition of tert-butyl radicals to 3 was estimated to be 2.3 × 106 M-1 s-1 at 60 °C. The overall ratio of addition to H abstraction in the reaction of 2 with 3 was 5:1.

Introduction The initiation process in free radical polymerization plays an important role in determining polymer properties such as stability, molecular weight distribution, and composition.1 In particular, initiator-derived end groups can have profound effects on the stability of the polymer toward thermal and/or photochemical degradation. The radical trapping technique developed by the CSIRO,2 employing the nitroxide 1 as radical scavenger,

has been used extensively to elucidate initiation mechanisms in free radical polymerization. Most early work involved the reaction of oxygen-centered radicals, that is, tert-butoxyl,2,3 benzoyloxyl,4,5 isopropoxycarbonyloxyl,5 cumyloxyl,6 hydroxyl,7 isopropoxyl,8 and ethoxyl radicals9 with monomer, since this †

Griffith University. CSIRO. NOF Corp. X Abstract published in AdVance ACS Abstracts, October 15, 1996. (1) (a) Moad, G.; Solomon, D. H. Aust. J. Chem. 1990, 43, 215-239. (b) Krstina, J.; Moad, G.; Solomon, D. H. Eur. Polym. J., 1989, 25, 767777. (2) Rizzardo, E.; Solomon, D. H. Polym. Bull. 1979, 1, 529-534. (3) (a) Griffiths, P. G.; Rizzardo, E.; Solomon, D. H. J. Macromol. Sci., Chem. 1982, 17, 45-50. (b) Busfield, W. K.; Jenkins, I. D.; Thang, S. H.; Rizzardo, E.; Solomon, D. H. J. Chem. Soc., Perkin Trans. 1 1988, 485490. (4) (a) Moad, G.; Rizzardo, E.; Solomon, D. H. J. Macromol. Sci., Chem. 1982, 17, 51-59. (b) Moad, G.; Rizzardo, E.; Solomon, D. H. Macromolecules 1982, 15, 909-914. (5) Cuthbertson, M. J.; Moad, G.; Rizzardo, E.; Solomon, D. H. Polym. Bull. 1982, 6, 647-651. (6) Rizzardo, E.; Serelis, A. K.; Solomon, D. H. Aust. J. Chem. 1982, 35, 2013-2024. (7) Grant, R. D.; Rizzardo, E.; Solomon, D. H. J. Chem. Soc., Perkin Trans. 2 1985, 379-384. (8) Busfield, W. K.; Jenkins, I. D.; Rizzardo, E.; Solomon, D. H.; Thang, S. H. J. Chem. Soc., Perkin Trans. 1 1991, 1351-1354.

technique is based on the fact that 1 reacts with carbon-centered radicals at almost diffusion-controlled rates10 but not with oxygen-centered radicals. However, there have been some observations of nitroxide trapped products arising from reactions of carbon-centered radicals (e.g., methyl radicals generated by β-scission of tertbutoxyl radicals, phenyl radicals11 from decarboxylation of benzoyloxyl radicals, and solvent-derived carbon-centered radicals12), with monomer. More recently, we have reported the reaction of “second generation” carbon-centered radicals,13 phosphorus-centered radicals,14 and sulfur-centered radicals15 with monomer followed by trapping by 1. Thus we have shown that the radical trapping technique can also be used to study reactions of radicals other than oxygen-centered ones with monomer. The work described in this paper is part of an ongoing investigation of the reaction of a combination of tert-alkoxyl radicals and alkyl radicals (generated by the thermolysis of tertalkyl peroxypivalates) with monomers. This paper reports the results of a study of the reaction of tert-butyl peroxypivalate (2) with methyl methacrylate (3) in the presence of the nitroxide 1. In the radical polymerization of acrylates and methacrylates, 2 is a widely used commercial initiator.16 However, no reports

‡ §

S0002-7863(96)02341-4 CCC: $12.00

(9) Busfield, W. K.; Jenkins, I. D.; Thang, S. H.; Rizzardo, E.; Solomon, D. H. Eur. Polym. J. 1993, 29, 397-400. (10) Beckwith, A. L. J.; Bowry, V. W.; Moad, G. J. Org. Chem. 1988, 53, 1632-1641. (11) Moad, G.; Rizzardo, E.; Solomon, D. H. Aust. J. Chem. 1983, 36, 1573-1588. (12) (a) Grant, R. D.; Griffiths, P. G.; Moad, G.; Rizzardo, E.; Solomon, D. H. Aust. J. Chem. 1983, 36, 2447-2454. (b) Bednarek, D.; Moad, G.; Rizzardo, E.; Solomon, D. H. Macromolecules 1988, 21, 1522-1528. (13) Bottle, S. E.; Busfield, W. K.; Heiland, K.; Jenkins, I. D.; Meutermans, W.; Monteiro, M. Progress in Pacific Polymer Science 3; Ghiggino, K. P., Ed.; Springer-Verlag: Berlin: Heidelberg, 1994; pp 8597. (14) Busfield, W. K.; Grice, I. D.; Jenkins, I. D. Aust. J. Chem. 1995, 48, 625-634. (15) (a) Busfield, W. K.; Heiland, K.; Jenkins, I. D. Tetrahedron Lett. 1994, 35, 6541-6542. (b) Busfield, W. K.; Heiland, K.; Jenkins, I. D. Tetrahedron Lett. 1995, 36, 1109-1112. (16) Sheppard, C. S. Encyclopedia of Polymer Science and Engineering, 2nd ed.; Klingsberg, A., Piccininni, R. M., Salvatore, A., Baldwin, T., Eds.; Wiley-Interscience: New York, 1988; Vol. 11, pp 1-21.

© 1996 American Chemical Society

Radical Polymerization of Methyl Methacrylate regarding the mechanism of initiation by 2 have been published.

J. Am. Chem. Soc., Vol. 118, No. 44, 1996 10825 Chart 1

Results and Discussion Initially the thermolysis of 2, both in the presence and in the absence of 1, was carried out in the absence of monomer in order to test for the possible induced decomposition of 2 by 1. It has been reported that 1 catalyzed the decomposition of certain organic peroxides (e.g., diacylperoxides5,17 and peroxydicarbonates5). The decomposition of 2 in cumene at 60 °C, measured by monitoring its disappearance by HPLC, satisfied first-order kinetics. The rate constant in the presence of 1, 2.66 × 10-5 s-1, was consistent with that in the absence of 1 (k ) 2.67 × 10-5 s-1) and also with the literature value, (k ) 2.70 × 10-5 s-1).18 This clearly indicates that 1 does not influence the thermal decomposition of 2. The initiator 2 is known to generate equimolar amounts of the two radicals via concerted two-bond scission in a solvent, such as cumene (Scheme 1).18,19 Scheme 1

Table 1. Ratios of Abstraction Products to the Tail Addition Product of tert-Butoxyl Radicals and Methyl Methacrylate (4) for Two Initiators at 60 °C relative product yields initiator

4

5

6

7

8

9

DTBPOa 2b

100 100

47.5 47.1

5.3 5.4

0.8 0.7

6.6 6.7

0.8 0.8

a DTBPO ) di-tert-butyl peroxyoxalate, [DTBPO] ) 0.020 M, [1] 0 0 ) 0.088 M, reaction time ) 1.25h. b [1]0 ) 0.080 M, [2]0 ) 0.040 M, reaction time ) 0.5h.

Scheme 2 Following the reaction of initiator 2 (0.040 M) with methyl methacrylate (3) as solvent in the presence of 1 (0.080 M) at 60 °C in vacuo, the reaction mixture was analyzed by HPLC and HPLC-MS. Alkoxyamines 4-11 (Chart 1) were formed in the relative percentage yields as shown. As expected, alkoxyamine 4 was the major product derived from tert-butoxyl radical addition to methyl methacrylate, 8 and 9 were minor products derived from methyl radical direct trapping and addition, respectively, and 10 and 11 were the products derived from competitive trapping and addition of tert-butyl radicals. Products 5, 6 and 7 are products formed by hydrogen abstraction from 3, followed by trapping (5 and 6) or further monomer tail addition and trapping (7). In a separate experiment, diacetyl peroxide was used as initiator to provide an exclusive source of methyl radicals. The only products formed were alkoxyamines 8 (78%) and 9 (22%). Thus 3 undergoes negligible H-abstraction by methyl radicals. In a further separate experiment, di-tert-butyl peroxyoxalate was used as initiator to provide a major source of tert-butoxyl radicals. Products 4-9 were all observed; the ratios of the abstraction products to the main tail addition product 4 are compared with the corresponding products from the reaction with pivalate 2 in Table 1. The close similarity of the ratios suggests that all the abstraction products from the pivalate reaction arise from tert-butoxyl radicals and that tert-butyl radicals result in negligible abstraction from 3. The reaction selectivity (preference for addition over abstraction) of both methyl and tert-butyl radicals is (17) Moad, G.; Rizzardo, E.; Solomon, D. H. Tetrahedron Lett. 1981, 22, 1165-1168. (18) The rate constant was calculated from the literature values at three temperatures (45, 55, and 65 °C). Fort, Jr., R. C.; Franklin, R. E. J. Am. Chem. Soc. 1968, 90, 5267-5268. (19) (a) Bartlett, P. D.; Simons, D. M. J. Am. Chem. Soc. 1960, 82, 17531756. (b) Koenig, T.; Wolf, R. J. Am. Chem. Soc. 1967, 89, 2948-2952. (c) Lorand, J. P.; Chodroff, S. D.; Wallace, R. W. J. Am. Chem. Soc. 1968, 90, 5266-5267. (d) Pryor, W. A.; Morkved, E. H.; Bickley, H. T. J. Org. Chem. 1972, 37, 1999-2005.

therefore similar to that of cyclohexyl radicals reported by Giese et al.20 The range and yields of products can be adequately explained by Scheme 2. The yield of products derived from tert-butyl radicals (49.3%) was slightly less than that derived from tertbutoxyl radicals (50.7%) presumably due to partial product decomposition. A series of experiments was carried out for various times in order to test the stability of products in the reaction system. Product 11 was a prime suspect whereas the tert-butyl adduct 10 is expected to be stable under the reaction conditions.10 A lower trap concentration was used to enhance the yield of 11, which, according to Scheme 2, is formed from addition of tert-butyl radicals to methyl methacrylate in (20) Giese, B.; Meixner, J. Chem. Ber. 1981, 114, 2138-2145.

10826 J. Am. Chem. Soc., Vol. 118, No. 44, 1996

Nakamura et al.

Table 2. Yields of All Products Relative to the Tail Addition Product of tert-Butoxyl Radicals and Methyl Methacrylate (4) at 60 °Ca relative product yields entry

[T]0/M

reaction time/h

4

5

6

7

8

9

10

11

4-9

10 + 11

1 2 3 4 5 6 7 8 9

0.040 0.040 0.040 0.040 0.040 0.060 0.080 0.120 0.160

0.5 1.0 2.0 3.0 5.0 0.5 0.5 0.5 0.5

100 100 100 100 100 100 100 100 100

47.1 46.9 47.2 47.7 48.5 47.0 47.1 47.1 47.2

5.0 5.0 5.0 5.0 4.7 5.3 5.4 5.6 5.8

1.3 1.4 1.5 1.6 1.8 0.9 0.7 0.5 0.3

5.9 6.0 6.0 6.0 5.9 6.5 6.7 6.9 7.1

1.5 1.4 1.6 1.6 2.0 1.1 0.8 0.6 0.4

104.2 101.1 98.1 90.6 84.9 118.0 128.3 136.7 142.6

46.6 40.4 28.9 22.3 11.9 35.7 28.0 20.3 15.3

160.8 160.7 161.3 161.9 162.9 160.8 160.7 160.8 160.8

150.8 141.5 127.0 112.9 96.8 153.7 156.3 157.0 157.9

a

[2]0 ) 0.040 M.

Scheme 3

competition with reaction with trap. Relative product yields are shown in Table 2. The results clearly indicate that product 11 is decomposing under the reaction conditions. A blank experiment was carried out in order to test the stability of 11. Thermolysis of 11 in methyl methacrylate in the presence of 1 and in the absence of 2 indicated the rapid decomposition of 11 (75% of 11 was decomposed after 2 h at 60 °C) to form hydroxyamine 15. This type of decomposition has been reported previously.21 The fact that compound 15 was not detected in the reaction products of 2 with methyl methacrylate may be due to oxidation of 15 by 2.21b

Methyl, tert-butyl, and acyloxymethyl radicals 14 all underwent competitive addition/trapping reactions (Scheme 3). A series of experiments was carried out for various initial concentrations of 1 in order to investigate the reactions of these alkyl radicals. The product yields, relative to the yield of the major addition product 4 (taken as 100), are summarized in Table 2. As expected, the relative amounts of tert-butoxyl radical-derived products was basically independent of the initial concentration of T ([T]0), that is, 4:5:6 + 7:8 + 9 ) 100:47.1: 6.1:7.6. On the other hand, tert-butyl radical-derived products were obtained in higher yield at higher [T]0 since the yield of 11 (and therefore the amount of the decomposition of 11) was reduced relative to the formation of 10. The ratios of products derived from competitive reactions of methyl, tert-butyl, and acyloxymethyl radicals 14, should be proportional to the ratio of the concentrations of T and MMA [see eq 1, where kadd and

kadd (R′-MMA-T) [T] ) kT [MMA] (R′-T)

(1)

kT are the general rate constants for the reaction of radical R′ with MMA and T respectively and (R′-T) and (R′-MMAT) are the yields of these compounds relative to the yield of 4]. (21) (a) Howard, J. A.; Tait, J. C. J. Org. Chem. 1978, 34, 4279-4283. (b) Grattan, D. W.; Carlsson, D. J.; Howard, J. A.; Wiles, D. M. Can. J. Chem. 1979, 57, 2834-2842.

Figure 1. Thermolysis of 2 in cumene in the presence (b) and in the absence of 1 (0) at 60 °C.

The total yield of alkyl radicals, (R′•), is given by eq 2. As the conversion of pivalate is very low (half-life of 2 at 60 °C is ca. 7 h in cumene), the amounts of MMA and T consumed are very small, and these concentrations can be assumed to be constant and equal to the initial values ([MMA]0 ) 9.1 M at 60 °C6 and [T]0 as shown in Table 2). Therefore, according to eq 3 (derived from eqs 1 and 2), the plot of 1/(R′-T) vs 1/[T]0 should be linear with 1/(R′•) as the intercept.

(R′•) ) (R′-T) + (R′-MMA-T)

(2)

kadd[MMA]0 1 1 1 ) + (R′-T) kT(R′•) [T]0 (R′•)

(3)

The linear relationships, shown in Figure 2, provide strong support for the validity of the experimental data and for Scheme 2. The intercepts give the total yield of radicals based on the yields of the directly trapped radical products, R′-T, and is independent of the yields of the products R′-MMA-T. This eliminates the problem of the decomposition of 11. The total yield of tert-butyl radicals is 162.9 (relative to 4 as 100), in very close agreement with the total yield of products from tertbutoxyl radicals (see Table 2). This shows that 2 generates equimolar amounts of the two radicals in monomer as well as in cumene.18,19 The yields of other alkyl radicals relative to alkoxyamine 4 were also confirmed [methyl (6.1) and acyloxymethyl radical 14 (7.6) respectively]. Assuming that tertbutyl radicals (the yield of 11) ) (the total yield of tert-butoxyl radical-derived products) - (the yield of 10), the ratios of kadd/ kT for methyl, tert-butyl, and acrylate radicals are obtained from the linear (R′-MMA-T)/(R′-T) vs 1/[T]0 plot according to eq 1 (shown in Figure 2). The resulting values were 1.1 × 10-3, 2.4 × 10-3, and 1.2 × 10-3, respectively, and if kT is assumed to have the same

Radical Polymerization of Methyl Methacrylate

Figure 2. 1/(R′-T) vs 1/[T]0 plots for alkyl radicals. (a) tert-Butyl radicals (0). (b) Methyl radicals (4) and acyloxymethyl radicals 14 (O).

J. Am. Chem. Soc., Vol. 118, No. 44, 1996 10827 Thus tert-butyl radicals add to methyl methacrylate about twice as fast as do methyl radicals.23 On the other hand, 14 shows about the same reactivity as methyl radicals. Probably the increased steric bulk of 14 (relative to methyl radicals) is compensated by an increased nucleophilicity resulting from the R-oxygen atom. Caronna et al.24 have reported that the relative reactivity of alkyl radicals in addition reactions to diethyl vinyl phosphonate is in the general order methyl < n-alkyl < secalkyl < tert-alkyl radicals. Also Baban et al.25 reported that isopropyl radicals react with acrylonitrile and methyl acrylate about 7 times faster than n-heptyl radicals at 0 °C. From these results, Giese26 concluded that for relatively small alkyl substituents on the carbon-centered radical, polar effects, which increase the nucleophilicity of the radical with an increasing number of alkyl groups at the radical center, are apparently more important than substituent effects on the stability of the radicals. The absolute rate constant for addition of tert-butyl radicals to methyl methacrylate, kadd(But•), was also estimated to be ca. 2.2 × 106 M-1 s-1 using the value10 of kT ) 9.1 × 108 M-1 s-1. This is consistent with a literature report27 that the addition of tert-butyl radicals to methyl methacrylate occurs >7.6 times faster than the corresponding addition to styrene (k ) 2.3 × 105 M-1 s-1 at 60 °C).28 Furthermore, the resulting value, kadd(CH3•) ) 1.0 × 106 M-1 s-1 at 60 °C, agrees with the value recently reported by Zytowski et al.,29 kadd(CH3•) ) 4.9 × 105 M-1 s-1 at 24 °C and Ea ) 16.0 kJ mol-1. In conclusion, in the reaction of 2 with 3, tert-butyl radicals and methyl radicals generated from the thermolysis of 2 underwent selective tail addition to 3, while tert-butoxyl radicals were responsible for all of the H-abstraction products. The rate constant for addition of tert-butyl radicals to 3 was about twice that for addition of methyl radicals and it was estimated to be 2.2 × 106 M-1 s-1 at 60 °C. About 83% of initiator-derived free radicals, that is, tert-butoxyl (31%), tert-butyl (50%), and methyl radicals (2%), underwent addition to 3. Experimental Section

Figure 3. (R′-MMA-T)/(R′-T) vs 1/[T]0 plots for alkyl radicals. tert-Butyl radicals (0), methyl radicals (4), and acyloxymethyl radicals 14 (O).

value for any carbon-centered radical,22 the relative reactivities of the three radicals toward tail addition can be estimated as follows:

kadd(CH3•):kadd(But•):kadd(14) ) 1.0:2.2:1.1 (22) Though the rate constants for nitroxide trapping of alkyl radicals by 1 were reported to be almost diffusion controlled (ca. 1 × 109 M-1 -1 s ),10 the values depend on the structure of the radicals trapped (e.g., steric protection of the radical center and resonance stability of the radical). Therefore, rate constants for radicals, 12 and 13 are likely to be lower than 109 M-1 s-1 (Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc., 1992, 114, 4992-4996).

HPLC was performed using Shimadzu LC-9A liquid chromatograph fitted with a Waters Nova-Pak C18, ODS column, connected to a Shimadzu UV spectrophotometric detector set at 270 nm and a CR6A computing integrator. Peak areas from HPLC chromatograms were converted directly into percent molar yields of products. Alkoxyamine compounds containing 1 mol equiv of the radical trapping moiety as UV chromophore have been shown to have almost identical molar extinction coefficients at 270 nm.6 NMR spectra were recorded on a Varian Gemini-200 (200MHz) spectrometer, using deuterated chloroform or methanol as solvent. Chemical shifts for 1H NMR spectra are relative to residual CHCl3 (δ 7.24 ppm) and for 13C NMR spectra are relative to the central peak of the triplet resonance due to CDCl3 (δ 77.0 ppm) or the central peak of seven-line multiplet resonance due to CD3OD (δ 49.0 ppm). HPLC-electrospray mass spectra were obtained with a Single Quadrupole VG Platform II mass spectrometer, coupled to a MassLynx data system. (23) If the trapping rate of methyl radicals is faster than that of tertbutyl radicals (9.1 × 108 M-1 s-1 at 18 °C),10 this ratio becomes smaller correspondingly. (24) Caronna, T.; Citterio, A.; Ghirardini, M.; Minisci, F. Tetrahedron 1977, 33, 793-796. (25) Bahan, J. A.; Roberts, B. P. J. Chem. Soc., Perkin Trans. 2 1981, 161-166. (26) Giese, B. Angew. Chem., Int. Ed. Engl. 1983, 22, 753-764. (27) Moad, G.; Solomon, D. H. ComprehensiVe Polymer Science; Eastmond, G. C., Ledwith, A., Russo, S., Sigwalt, P., Eds.; Pergamon Press: Oxford, 1989; Vol. 3, p 108. (28) The value was calculated from the Arrhenius equation (log A ) 7.6 ( 0.2 M-1 s-1, Ea ) 14.3 ( 1.0 kJ mol-1). Mu¨nger, K.; Fischer, H. Int. J. Chem. Kinet. 1985, 17, 809-829. (29) Zytowski, T.; Fischer, H. J. Am. Chem. Soc. 1996, 118, 437-439.

10828 J. Am. Chem. Soc., Vol. 118, No. 44, 1996 Materials. Methyl methacrylate 3 was washed with 5% NaOH, dried over anhydrous Na2SO4, and distilled at atmospheric pressure. Cumene was washed with concentrated H2SO4, dried over anhydrous Na2SO4, and distilled at reduced pressure. Both solvents were stored in a refrigerator (-20 °C). tert-Butyl peroxypivalate (2) was prepared by the reaction of pivaloyl chloride with tert-butyl hydroperoxide in alkaline solution. 2 was 99.7% pure (iodometric titration30); νmax(neat)/ cm-1 1769vs (CdO); m/z 197 (M + Na)+, 175 (M + H)+. Acetyl peroxide was prepared by the reaction of acetic anhydride with hydrogen peroxide in the presence of sodium carbonate.31 Di-tert-butyl peroxyoxalate,32 nitroxide 1,33 and hydroxyamine 158 were prepared by literature procedures. Kinetic Experiment. Thermolysis of 2 in the Presence of 1 in Cumene. Cumene solutions (5 mL) containing 2 (0.05 M) and 1 (0.11 M) were placed in glass ampules, which were purged with nitrogen, sealed, and immersed in a constant temperature bath regulated at 60 °C. At successive time intervals, ampules were removed from the bath. The concentration of peroxide in each solution was measured by HPLC (at 250 nm). The reaction was followed up to 50% decomposition and exhibited first-order kinetics. Trapping Experiments. Typical Procedure for Reaction of 2 with 3 in the presence of 1. A solution of 2 and 1 in freshly distilled 3 was degassed by repeated freezing and thawing (three cycles) on a vacuum line (